Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 1
Slide 1: Introduction
Hello and welcome to Module 2 – Fundamentals
of Epidemiology. This presentation will focus on
issues of interpretation in epidemiologic studies.
My name is Jeffrey Bethel, Assistant Professor of
Epidemiology at East Carolina University, Brody
School of Medicine
Slide 2: Acknowledgements
Slide 3: Presentation Objectives
There are four objectives for this presentation.
They are to describe the key features of
information and selection bias; identify the ways
selection and information bias can be minimized
or avoided; Implement the methods for
assessing and controlling confounding; and
identify uses of the Surgeon General’s Guidelines
for establishing causality.
Developed through the APTR Initiative to Enhance Prevention and Population
Health Education in collaboration with the Brody School of Medicine at East
Carolina University with funding from the Centers for Disease Control and
Prevention
Issues of Interpretation in Epidemiologic Studies
Developed through the APTR Initiative to Enhance Prevention and Population
Health Education in collaboration with the Brody School of Medicine at East
Carolina University with funding from the Centers for Disease Control and
Prevention
Issues of Interpretation in Epidemiologic Studies
APTR wishes to acknowledge the following individual that developed this module:
Jeffrey Bethel, PhDDepartment of Public Health
Brody School of Medicine at East Carolina University
This education module is made possible through the Centers for Disease Control and Prevention (CDC) and the
Association for Prevention Teaching and Research (APTR) Cooperative Agreement, No. 5U50CD300860. The module
represents the opinions of the author(s) and does not necessarily represent the views of the Centers for Disease
Control and Prevention or the Association for Prevention Teaching and Research.
APTR wishes to acknowledge the following individual that developed this module:
Jeffrey Bethel, PhDDepartment of Public Health
Brody School of Medicine at East Carolina University
This education module is made possible through the Centers for Disease Control and Prevention (CDC) and the
Association for Prevention Teaching and Research (APTR) Cooperative Agreement, No. 5U50CD300860. The module
represents the opinions of the author(s) and does not necessarily represent the views of the Centers for Disease
Control and Prevention or the Association for Prevention Teaching and Research.
1. Describe the key features of selection and information bias
2. Identify the ways selection and information bias can be minimized or avoided
3. Implement the methods for assessing and controlling confounding
4. Identify uses of the Surgeon General’s Guidelines for establishing causality
1. Describe the key features of selection and information bias
2. Identify the ways selection and information bias can be minimized or avoided
3. Implement the methods for assessing and controlling confounding
4. Identify uses of the Surgeon General’s Guidelines for establishing causality
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 2
Slide 4: Hierarchy of Study Designs
The goal of public health as well as clinical
medicine is to modify the natural history of
disease, to decrease morbidity and mortality. So
how do researchers and public health
practitioners do this? They carry out formal
studies to identify risk factors for various
diseases or other health outcomes. So here is an
informative figure showing the hierarchy of various epidemiologic study designs. The
goal of all these studies is the same and that is to identify exposures or characteristics
that are associated with disease or other health related outcomes.
Slide 5: Linking Exposure to Outcome
Here is a figure depicting this goal of linking
exposures to outcomes. The result of these
formal epidemiologic studies is observed
associations. That is associations between
exposures or characteristics and diseases or
outcomes.
Slide 6: Linking Exposure to Outcome
After the study is carried out, and a disease is
linked to an exposure or outcome, you need to
ask if the observed association is biased
confounding or causal. We’re going to go
through each one of these items individually.
And we’re going to start with bias first.
Smith, AH. The Epidemiologic Research Sequence. 1984Smith, AH. The Epidemiologic Research Sequence. 1984
Exposure or Characteristic
Disease orOutcome
Observed Association
Exposure or Characteristic
Disease orOutcome
Observed Association
Exposure or Characteristic
Disease orOutcome
Observed Association
Is it: biased, confounded,
or causal?
Exposure or Characteristic
Disease orOutcome
Observed Association
Is it: biased, confounded,
or causal?
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 3
Slide 7: Bias
So bias is any systematic error in the design,
conduct, or analysis of a study that results in a
mistaken estimate of an exposure’s effect on the
risk of disease.
Slide 8: Bias
Now these errors or bias can make it appear as if
an exposure is associated with an outcome when
the exposure and the outcome are really not
associated. That is also thought of as biasing
away from the null. The errors in the bias can
also mask an association where there really is an
association. And that is called biasing towards
the null. There really is an association however a bias estimate makes it appear as
though there is no association. As we talked about in the definition of bias, bias is
primarily introduced by the investigators or participants, in the design, conduct, or
analysis of the study. And we’re going to talk about two forms of bias in a study;
selection and information bias.
Slide 9: Selection Bias
Selection bias occurs when the methods that
investigators use to identify study participants
results in a mistaken estimate. That is the
estimate that investigators generate is different
than what would have been obtained from the
entire population targeted for the study. Again
this gets back to a systematic error made by the
investigators. Specifically the error was made in selecting one or more of the study
groups that are being compared in the epidemiologic study.
Any systematic error in the design, conduct, or analysis of a study that results in a mistaken estimate of an exposure’s effect on the risk of disease
Any systematic error in the design, conduct, or analysis of a study that results in a mistaken estimate of an exposure’s effect on the risk of disease
Can create spurious association when there really is not one (bias away from the null)
Can mask an association when there really is one (bias towards the null)
Primarily introduced by the investigators or study participants
Selection and information bias
Can create spurious association when there really is not one (bias away from the null)
Can mask an association when there really is one (bias towards the null)
Primarily introduced by the investigators or study participants
Selection and information bias
Results from procedures used to study participants that lead to a result different from what would have been obtained from the entire population targeted for study
Systematic error made in selecting one or more of the study groups to be compared
Results from procedures used to study participants that lead to a result different from what would have been obtained from the entire population targeted for study
Systematic error made in selecting one or more of the study groups to be compared
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 4
Slide 10: Selection Bias Example
And we’re going to talk about four types of
selection bias. Control selection bias, self-
selection bias, differential referral, surveillance,
or diagnosis bias, and loss to follow-up.
Slide 11: Selection Bias in a Case-Control Study
Now here is an example in which selection bias
may occur in a case-control study. Specifically
control selection bias. So the research question
was, do Pap smears prevent cervical cancer? So
in this study cases were diagnosed at a city
hospital. Controls were randomly sampled from
households in the same city by canvassing the
neighborhood on foot. So here was the observed association. Here are the data from
that hypothetical study. Women who were cases and controls, and women who
received Pap smears and no Pap smears. So 250 cases, 250 controls, 100 of the cases
received a Pap smear and 100 of the controls received a Pap smear. The resulting odds
ratio was 1.0. That is there is no association between Pap smears and risk of cervical
cancer.
Slide 12: Selection Bias in a Case-Control Study
However, cases came from the hospital and
controls from the neighborhood around the
hospital. So here is where the bias comes in.
Only controls who were at home during
recruitment for the study were actually involved
in the study. Now these controls who were
home while the neighborhoods were being
canvassed were less likely to work and less likely to have regular checkups and Pap
smears. Therefore, inclusion in the study as a control was not independent of the
exposure. That is that being included in the study as a control was dependent on the
exposure, in that case Pap smear.
Control selection bias
Self-selection bias
Differential referral, surveillance, or diagnosis bias
Loss to follow-up
Control selection bias
Self-selection bias
Differential referral, surveillance, or diagnosis bias
Loss to follow-up
Cases Controls
Pap Smear 100 100
No PapSmear
150 150
Total 250 250
Question: Do Pap smears prevent cervical cancer? Cases diagnosed at a city hospital. Controls randomly sampled from household in same city by canvassing the neighborhood on foot. Here is the observed relationship:
OR = (100 x 150) / (100 x 50) = 1.0 There is no association between Pap smears and risk of cervical cancer (40% of cases and 40% of controls had Pap smears)
Cases Controls
Pap Smear 100 100
No PapSmear
150 150
Total 250 250
Question: Do Pap smears prevent cervical cancer? Cases diagnosed at a city hospital. Controls randomly sampled from household in same city by canvassing the neighborhood on foot. Here is the observed relationship:
OR = (100 x 150) / (100 x 50) = 1.0 There is no association between Pap smears and risk of cervical cancer (40% of cases and 40% of controls had Pap smears)
Recall: Cases from the hospital and controls come from the neighborhood around the hospital
Now for the bias: Only controls who were at home during recruitment for the study were actually included in the study. Women at home were less likely to work and less likely to have regular checkups and Pap smears. Therefore, being included in the study as a control is not independent of the exposure.
Recall: Cases from the hospital and controls come from the neighborhood around the hospital
Now for the bias: Only controls who were at home during recruitment for the study were actually included in the study. Women at home were less likely to work and less likely to have regular checkups and Pap smears. Therefore, being included in the study as a control is not independent of the exposure.
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 5
Slide 13: Selection Bias in a Case-Control Study
So here are the data for the true relationship.
Again, 250 cases, 250 controls. 100 of the cases
received a Pap smear, 150 of the controls
received a Pap smear. The resulting odds ratio is
0.44. That is there is a 56% reduced risk of
cervical cancer among women who received a
Pap smear. This is the true relationship. This is
the unbiased estimate. The odds ratio of 1.0 was the biased estimate.
Slide 14: Selection Bias in a Case-Control Study
Another form of selection bias is self-selection
bias. This occurs when refusal or non-response
to being included in the study is related to both
the exposure and the disease. For example, if
exposed cases were more or less likely to
participate than participants in the other
categories that can result in self-selection bias.
The people, either cases or controls, were more or less likely to participate. Now the
best way to avoid self-selection bias is to maintain high participation rates in all groups or
categories of participants in the study.
Slide 15: Selection Bias in a Case-Control Study
Now here is differential surveillance, diagnosis or
referral bias. So here is an example related to
exposure. In a case-control study looking at
venous thromboembolism (VT) and oral
contraceptive (OC) use, cases were 20-44 year
old women hospitalized for VT. Controls were
the same age group hospitalized for acute illness
or elective surgery at the same hospitals. In this instance hospital controls were used.
The resulting odds ratio from this study was 10.2. That is OC use was highly associated
with venous thromboembolism.
Cases Controls
Pap Smear 100 150
No PapSmear
150 100
Total 250 250
Question: Do Pap smears prevent cervical cancer? Cases diagnosed at a city hospital. Controls randomly sampled from household in same city by canvassing the neighborhood on foot. Here is the true relationship:
OR = (100)(100) / (150)(150) = .44 56% reduced risk of cervical cancer among women who had Pap smears as compared to women who did not (40% of cases had Pap smears versus 60% of controls)
Cases Controls
Pap Smear 100 150
No PapSmear
150 100
Total 250 250
Question: Do Pap smears prevent cervical cancer? Cases diagnosed at a city hospital. Controls randomly sampled from household in same city by canvassing the neighborhood on foot. Here is the true relationship:
OR = (100)(100) / (150)(150) = .44 56% reduced risk of cervical cancer among women who had Pap smears as compared to women who did not (40% of cases had Pap smears versus 60% of controls)
Refusal or nonresponse by participants that is related to both exposure and disease
e.g. if exposed cases are more/less likely to participate than participants in other categories
Best way to avoid is to obtain high participation rates
Refusal or nonresponse by participants that is related to both exposure and disease
e.g. if exposed cases are more/less likely to participate than participants in other categories
Best way to avoid is to obtain high participation rates
Example related to exposure
CC study: venous thromboembolism (VT) and oral contraceptive (OC) use
Cases: 20-44 yo, hospitalized for VT
Controls: 20-44 yo, hospitalized for acute illness or elective surgery at same hospitals
Result: OR = 10.2
Example related to exposure
CC study: venous thromboembolism (VT) and oral contraceptive (OC) use
Cases: 20-44 yo, hospitalized for VT
Controls: 20-44 yo, hospitalized for acute illness or elective surgery at same hospitals
Result: OR = 10.2
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 6
Slide 16: Selection Bias in a Case-Control Study
However, authors acknowledged that this high
odds ratio might be due to the bias in criteria for
hospital admission. That is who was admitted.
So previous studies had liked VT to OC use.
Therefore there was a tendency to hospitalize
patients based on exposure status of oral
contraceptive use and that led to a stronger
association between VT and OC than what truly existed.
Slide 17: Selection Bias in a Cohort Study
Here’s an example in the form of selection bias in
a cohort study that is loss to follow-up. One
study compared HIV incidence rates among IV
drug users (IVDU) in New York City during a six
year period using 10 incidence studies. The HIV
rates ranged from 0 to 2.96 per 100 person-
years. And these were well below the incidence
rates in NYC from the late 70s and early 80s to mid-80s and the early 90s. Now this study
resulted in funding cuts to drug treatment and prevention programs because this study
showed that HIV incidence rates were much lower than they were in the 70s and 80s.
Slide 18: Selection in a Cohort Study
Now the question is was this decline in HIV
incidence rates real? Closer inspection of these
10 individual cohort studies found that follow-up
rates ranged from 36% to 95% and only 2 of the
10 had follow-up rates greater than 80%. As we
know, loss to follow-up can introduce bias into
the study. In this case only 2 of the 10 had a high
follow-up rate. Also, sample sizes ranged from 96 to 1,671 in these 10 studies. Now the
solution to loss to follow-up is to simply minimize loss to follow-up. That involves
implementing methods and procedures in your study design to minimize loss to follow-
up. You could maintain accurate contact information for the participants, and contact
information for people who will know where your participants are so as you do not lose
them.
Authors acknowledged high OR might be due to “bias in the criteria for hospital admission”
Previous studies linked VT to OC
Health care provider more likely to hospitalize women with VT symptoms who were taking OC than symptomatic women who were not taking OC
Authors acknowledged high OR might be due to “bias in the criteria for hospital admission”
Previous studies linked VT to OC
Health care provider more likely to hospitalize women with VT symptoms who were taking OC than symptomatic women who were not taking OC
Compared HIV incidence rates among IVDU in NYC from 1992-97 through 10 incidence studies to previous years
HIV incidence rates (IR) range from 0 to 2.96 per 100 person-years (py)
Well below IR in NYC from late 70s and early 80s (13 per 100 py) to mid 80s and early 90s (4.4 per 100 py)
Resulted in funding cuts to drug treatment and prevention programs
Compared HIV incidence rates among IVDU in NYC from 1992-97 through 10 incidence studies to previous years
HIV incidence rates (IR) range from 0 to 2.96 per 100 person-years (py)
Well below IR in NYC from late 70s and early 80s (13 per 100 py) to mid 80s and early 90s (4.4 per 100 py)
Resulted in funding cuts to drug treatment and prevention programs
Was decline real?
Follow-up rates in 10 cohorts ranged from 36% to 95%
Only 2 reported >80%
Sample size ranged from 96 to 1,671
Solution: minimize loss to follow-up
Was decline real?
Follow-up rates in 10 cohorts ranged from 36% to 95%
Only 2 reported >80%
Sample size ranged from 96 to 1,671
Solution: minimize loss to follow-up
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 7
Slide 19: Selection Bias – Can We Fix It?
So selection bias, can we fix it? Can we do
anything about it? The simple answer is no, you
simply need to put in measures into your design
and conduct the study to avoid it in the first
place. For example you need to use the same
criteria for selecting cases and controls. Obtain
all relevant participant records. That is, maintain
complete data on everyone. Obtain high participation rates. That is, minimize loss to
follow-up. And you need to take into account diagnostic and referral patterns of the
disease of interest, or the diseases that you are using for your hospital controls.
Slide 20: Information Bias
The second major type of bias we’re going to talk
about is information bias. That arises when from
a systematic difference in the way that exposure
or outcome is measured between various groups
under investigation. Just like selection bias,
information bias can bias your results towards or
away from the null. That is, mask an apparent
association or create a spurious one. Information bias can occur in prospective as well as
retrospective studies. And the two types we’re going to talk about include recall and
interviewer bias.
Slide 21: Recall Bias in a Case-Control Study
Here’s an example of recall bias in a case-control
study. In this example it is a case control study
examining birth defects. Controls in this study
were healthy infants. Cases were infants with
birth defects. Exposure data was collected at
postpartum interviews with the infants’ mothers.
Now in this situation there could be a differential
level in the accuracy of the information provided by the cases and controls which could
result in an over or underestimate of the measure of association. So the data collected
from the cases could be more accurate than the data collected from the controls. That is
the mothers of the cases have been pouring over their brains for the history of various
exposures during pregnancy which may have contributed to their infants being
malformed. Whereas controls, they have healthy infants, they haven’t really spent a lot
of time reviewing their history during pregnancy. Or conversely, cases may underreport
No – need to avoid it when you design and conduct the study
For example
Use the same criteria for selecting cases and controls
Obtaining all relevant participant records
Obtaining high participation rates
Taking in account diagnostic and referral patterns of disease
No – need to avoid it when you design and conduct the study
For example
Use the same criteria for selecting cases and controls
Obtaining all relevant participant records
Obtaining high participation rates
Taking in account diagnostic and referral patterns of disease
Arises from a systematic difference in the way that exposure or outcome is measured between groups
Can bias towards or away from the null
Occurs in prospective and retrospective studies
Includes recall bias and interviewer bias
Arises from a systematic difference in the way that exposure or outcome is measured between groups
Can bias towards or away from the null
Occurs in prospective and retrospective studies
Includes recall bias and interviewer bias
Case-control study of birth defects
Controls: healthy infants
Cases: malformed infants
Exposure data collected at postpartum interviews with infants’ mothers
Controls or cases may have underreported exposure, depending on nature of exposure
Case-control study of birth defects
Controls: healthy infants
Cases: malformed infants
Exposure data collected at postpartum interviews with infants’ mothers
Controls or cases may have underreported exposure, depending on nature of exposure
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 8
their history particularly if it is a socially sensitive exposure such as alcohol consumption
during pregnancy.
Slide 22: Methods to Minimize Recall Bias
Now what are the methods to minimize recall
bias? In the context of a case-control study you
could use a diseased control group. So as the
participants in the diseased control group have
some disease unrelated to the outcome of
interest could also be pouring over their history of
exposures. You could use and design a structured
questionnaire. You could use a self-administered questionnaire so as to make it easier
for participants to recall sensitive exposures such as consuming alcohol or smoking
during pregnancy. You could bypass the participant in terms of self-report and use
biological measurements. It is tough to get around socially sensitive exposures. And
lastly you could mask the participants to the study hypotheses. So as to illicit more
accurate information from the study participants.
Slide 23: Interviewer Bias
The second type of information bias we’re going
to talk about is interviewer bias. That results
from a systematic difference in soliciting,
recording, and interpreting of data collected
during a study. In a case-control study this could
occur when exposure information is sought or
collected differently between cases and controls.
In a cohort study, outcome or disease information is collected differently between the
exposed and unexposed individuals. There is a simple solution to this and that is to
simply mask the interviewers or the people collecting the data as to the disease status or
exposure status of the participants. You can also use standardized questionnaires or
standardized methods of outcome or exposure ascertainment.
Select diseased control group
Design structured questionnaire
Use self-administered questionnaire
Use biological measurements
Mask participants to study hypotheses
Select diseased control group
Design structured questionnaire
Use self-administered questionnaire
Use biological measurements
Mask participants to study hypotheses
Systematic difference in soliciting, recording, interpreting information
Case-control study: exposure information is sought when outcome is known
Cohort study: outcome information is sought when exposure is known
Solutions: Mask interviewers, use standardized questionnaires or standardized methods of outcome (or exposure) ascertainment
Systematic difference in soliciting, recording, interpreting information
Case-control study: exposure information is sought when outcome is known
Cohort study: outcome information is sought when exposure is known
Solutions: Mask interviewers, use standardized questionnaires or standardized methods of outcome (or exposure) ascertainment
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 9
Slide 24: Association
So that covers the examination of whether our
observed association is biased or not. The next
question we’re going to ask about our observed
association is whether this association is
confounded or not.
Slide 25: Confounding
Here is a diagram of a possibly confounded
relationship between A and B, confounded by
the variable X. Confounding is really a mixing of
effects. That is the association between
exposure and the disease of interest is distorted
by the effect of a third variable that is also
associated with the disease as well as the
exposure. You can also think about confounding as an alternate explanation between A
and B in this diagram. That is an alternate explanation of the observed association
between the exposure of interest A and the disease of interest B.
Slide 26: Criteria for Confounding
There are three criteria for determining whether
a variable is confounding the relationship
between A and B. So in order for this factor or
variable X to be a confounder, all of the following
must be true. This factor must be associated
with the Disease B. That is it’s a risk factor or a
preventive factor. This factor must also be
associated with Factor A. That is the exposure. Finally this factor must not be a result of
Factor A. That is, it must not lie on the causal pathway from A to B. Or it may not be a
result of A.
Exposure or Characteristic
Disease orOutcome
Observed Association
Is it: biased, confounded,
or causal?
Exposure or Characteristic
Disease orOutcome
Observed Association
Is it: biased, confounded,
or causal?
X
A B A mixing of effects – association between exposure and disease is distorted by the effect of a third variable that is associated with the disease
Alternate explanation for observed association between an exposure and disease
X
A B A mixing of effects – association between exposure and disease is distorted by the effect of a third variable that is associated with the disease
Alternate explanation for observed association between an exposure and disease
X
A B In order for a factor (X) to be a confounder, all of the following must be TRUE:
Factor X is associated with Disease B (risk factor or preventive factor)
Factor X is associated with Factor A (exposure)
Factor X is not a result of Factor A (not on causal pathway)
X
A B In order for a factor (X) to be a confounder, all of the following must be TRUE:
Factor X is associated with Disease B (risk factor or preventive factor)
Factor X is associated with Factor A (exposure)
Factor X is not a result of Factor A (not on causal pathway)
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 10
Slide 27: Example of Confounding
Here’s an example of a confounding relationship
in a previous study that was trying to link coffee
consumption with pancreatic cancer. And they
thought that perhaps smoking was confounding
the relationship between coffee consumption
and pancreatic cancer. So they went through the
three criteria to see if smoking was a
confounder. Criterion 1: is this confounding variable, smoking, associated with the
disease of interest or outcome, pancreatic cancer? And they found in their data that yes,
it was. Smoking was a risk factor for pancreatic cancer. Next, was smoking associated
with the exposure or coffee consumption? And they found in their data that yes, people
consuming coffee were more likely to also smoke. And looking outside their data, they
knew that smoking was not a result of coffee drinking. That smoking is not on the causal
pathway from coffee consumption to pancreatic cancer. All three criteria were satisfied
therefore they concluded that smoking was indeed confounding the relationship
between coffee consumption and pancreatic cancer. It was an alternate explanation
between coffee consumption and pancreatic cancer.
Slide 28: Impact of Confounding So what is the impact of confounding? It is similar to bias. You can pull the association away from the true association in either direction. So in positive confounding the confounded estimate or the confounded variable exaggerates the true association. So in the context of a cohort study, the true relative risk is 1. That is the exposure is not associated with the outcome. However, the
confounded relative risk could be 2. It is exaggerating the true association. There is also negative confounding when it hides the true association. For example, in a cohort study the true relative risk could be 2. That is exposure is related to the outcome. However, due to confounding, the relative risk is 1. There is no association between the exposure of interest and the disease. Again, it is similar to bias.
Pulls the observed association away from the true association
Positive confounding
Exaggerates the true association
True relative risk (RR) = 1.0 and confounded RR = 2.0
Negative confounding
Hides the true association
True RR = 2.0 and confounded RR = 1.0
Pulls the observed association away from the true association
Positive confounding
Exaggerates the true association
True relative risk (RR) = 1.0 and confounded RR = 2.0
Negative confounding
Hides the true association
True RR = 2.0 and confounded RR = 1.0
Smoking is associated with pancreatic cancer
Smoking is associated with coffee drinking
Smoking is not a result of coffee drinking
CoffeeConsumption
PancreaticCancer
Smoking
Smoking is associated with pancreatic cancer
Smoking is associated with coffee drinking
Smoking is not a result of coffee drinking
CoffeeConsumption
PancreaticCancer
Smoking
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 11
Slide 29: Hypothetical Cohort Study of Obesity and Dementia So now let’s work through a hypothetical cohort study examining the association between obesity and dementia. So a hypothetical study was conducted among 2,000 people looking at obesity and dementia. Here are the results of the data. We calculated the relative risk of 4.0. That is the incidence of dementia among the
obese participants was 4 times the incidence of dementia in the non-obese participants. So obesity was associated with dementia.
Slide 30: Association Between Obesity and
Dementia
So now we should ask ourselves, was age
confounding the association between obesity and
dementia?
Slide 31: Dementia Diabetes Cohort Study So we can look at our data and look at Criterion 1 first: Is age associated with our outcome, dementia? We can summarize our data in a 2x2 table and calculate relative risk and we get 4.0. That is age, and in this case older age, was associated with dementia.
Dementia No Dementia TOTAL
Obese 400 600 1,000
Not Obese 100 900 1,000
TOTAL 500 1,500 2,000
Relative Risk = = 4.0 (crude measure)400/1,000
100/1,000
Dementia No Dementia TOTAL
Obese 400 600 1,000
Not Obese 100 900 1,000
TOTAL 500 1,500 2,000
Relative Risk = = 4.0 (crude measure)400/1,000
100/1,000
Dementia TOTAL
Age Yes No
80-99Years
400 600 1,000
45-79Years
100 900 1,000
TOTAL 1,000 1,000 2,000
Relative Risk = = 4.0 400/1,000
100/1,000
Dementia TOTAL
Age Yes No
80-99Years
400 600 1,000
45-79Years
100 900 1,000
TOTAL 1,000 1,000 2,000
Relative Risk = = 4.0 400/1,000
100/1,000
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 12
Slide 32: Dementia and Diabetes Cohort Study
Criterion 2: Was age, the potential confounder,
associated with our outcome obesity? Here are
the resulting data. In examination of this
association we found out that yes, age is highly
associated with obesity.
Slide 33: Were Criteria for Confounding
Satisfied?
So were the criteria for confounding satisfied?
Criterion 1: Age was associated with dementia.
Criterion 2: Age was associated with obesity.
Age is not a result of obesity. Again that is not
from the data in the study. All 3 criteria were
satisfied. Therefore, yes age was confounding
the observed association between obesity and dementia. It just so happens that the
obese participants in the study were older and that was what accounted for the
observed association with dementia
Slide 34: Controlling Confounding
So really, confounding variables are nuisance
variables that get in the way of the relationship
that we really want to study. So how do we
control for it? We control for confounding in the
design and analysis phases. In design phase we
can group or individually match on the suspected
confounding factor. This is specifically regarding
a case-control study where we can match cases and controls on a variable such as age to
make them similar so they do not differ. In the analysis we can stratify our analyses
based on this confounding variable. Record our data stratified by age for example, young
and old. We can standardize our data. Or what’s most commonly done is conduct
multivariate analysis to adjust for the confounding variable.
Obese TOTAL
Age Yes No
80-99Years
900 100 1,000
45-79 Years
100 900 1,000
TOTAL 1,000 1,000 2,000
Relative Risk = = 9.0 900/1000
100/1000Odds ratio = = 81.0 900 x 900
100 x 100
Obese TOTAL
Age Yes No
80-99Years
900 100 1,000
45-79 Years
100 900 1,000
TOTAL 1,000 1,000 2,000
Relative Risk = = 9.0 900/1000
100/1000Odds ratio = = 81.0 900 x 900
100 x 100
Age
Obesity Dementia
Age is associated with dementia (RR=4.0)
Age is associated with obesity (OR=81.0)
Age is not a result of obesity (not from data)
Age
Obesity Dementia
Age is associated with dementia (RR=4.0)
Age is associated with obesity (OR=81.0)
Age is not a result of obesity (not from data)
Design phase
Group or individual matching on the suspected confounding factor
e.g. matching on age in case-control study
Analysis phase
Stratification
Standardization
Adjustment (multivariate analysis)
Design phase
Group or individual matching on the suspected confounding factor
e.g. matching on age in case-control study
Analysis phase
Stratification
Standardization
Adjustment (multivariate analysis)
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 13
Slide 35: Thoughts on Confounding
Some final thoughts on confounding. Again
confounding variables are these nuisance
variables that get in the way of examining the
relationship that we really want to study. Unlike
bias, it is not an error in the study. So these
confounding relationships are really valid
findings of relationships between factors and a
disease. However, failure to take into account confounding or to control for confounding
is an error and will bias the results of your study.
Slide 36: Association Between Exposure and
Outcome
So we’ve assessed whether the observed
association is biased or confounded. Let’s finally
look at whether our observed association
between our exposure and our outcome is
causal.
Slide 37: Epidemiologic Reasoning
First let’s talk a little about the epidemiologic
reasoning. The primary goal is to determine
whether there is a statistical association between
our exposure or characteristic and our disease.
We assess that through various studies. We can
study group characteristics in an ecological
study, or we can study individual characteristics
in a case-control, cross-sectional or cohort study. The key here is deriving inferences
regarding possible causal relationships using pre-determined criteria or guidelines. Again
we have this observed association, now we need to infer whether this association is
causal or not and we can use pre-determined guidelines to do this.
Not an error in the study
Valid finding of relationships between factors and disease
Failure to take into account confounding IS an error and can bias the results!
Not an error in the study
Valid finding of relationships between factors and disease
Failure to take into account confounding IS an error and can bias the results!
Exposure or Characteristic
Disease orOutcome
Observed Association
Is it: biased, confounded,
or causal?
Exposure or Characteristic
Disease orOutcome
Observed Association
Is it: biased, confounded,
or causal?
Determine whether a statistical association exists between characteristics or exposures and disease
Study of group characteristics (ecologic studies)
Study of individual characteristics (case-control and cohort studies)
Derive inferences regarding possible causal relationship using pre-determined criteria or guidelines
Determine whether a statistical association exists between characteristics or exposures and disease
Study of group characteristics (ecologic studies)
Study of individual characteristics (case-control and cohort studies)
Derive inferences regarding possible causal relationship using pre-determined criteria or guidelines
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 14
Slide 38: Causation
However, association between an exposure and
an outcome is not equal to causation. Consider
the following statement: If the rooster crows at
the break of dawn every morning, then the
rooster caused the sun to rise. Again these are
associated but they’re not causal. Causation
implies there is a true mechanism from the
exposure to the disease. More than just an observed association.
Slide 39: Koch-Henle Postulates (1880s)
So a little history lesson on causation. In this
instance in the context of infectious disease.
Jacob Henle and his student Robert Koch in the
1880s built a set of postulates on disease
causation based on the germ theory. There were
3 postulates. 1: the organism is always found
with the disease. This is also known as the
regular postulate. Second: the organism is not found with any other disease. That is, it is
exclusive. Third, the organism, isolated from one who has the disease, and cultured
through several generations, produces the disease. That is, it is reproducible in
experimental animals.
Slide 40: Koch-Henle Postulates (1880s)
Koch added that even when an infectious disease
cannot be transmitted to animals, (the third
postulate) the ‘regular’ and ‘exclusive’ presence
of the organism (the first two postulates) can
prove a causal relationship. However, unknown
at the time of these postulates in the mid 1800s,
was the idea of a carrier state, the idea of
asymptomatic infection, of multifactorial causation, or even the biologic spectrum of
disease. Those were not considered at this time. So causation needed to mature from
this idea of these three postulates.
Association is not equal to causation
Consider the following statement: If the rooster crows at the break of dawn, then the rooster caused the sun to rise
Causation implies there is a true mechanism from exposure to disease
Association is not equal to causation
Consider the following statement: If the rooster crows at the break of dawn, then the rooster caused the sun to rise
Causation implies there is a true mechanism from exposure to disease
1. The organism is always found with the disease (regular)
2. The organism is not found with any other disease (exclusive)
3. The organism, isolated from one who has the disease, and cultured through several generations, produces the disease (in experimental animals)
1. The organism is always found with the disease (regular)
2. The organism is not found with any other disease (exclusive)
3. The organism, isolated from one who has the disease, and cultured through several generations, produces the disease (in experimental animals)
Koch added that “Even when an infectious disease cannot be transmitted to animals, the ‘regular’ and ‘exclusive’ presence of the organism [postulates 1 and 2] proves a causal relationship
Unknown at the time of Koch-Henle (1840-1880) Carrier state
Asymptomatic infection
Multifactorial causation
Biologic spectrum of disease
Koch added that “Even when an infectious disease cannot be transmitted to animals, the ‘regular’ and ‘exclusive’ presence of the organism [postulates 1 and 2] proves a causal relationship
Unknown at the time of Koch-Henle (1840-1880) Carrier state
Asymptomatic infection
Multifactorial causation
Biologic spectrum of disease
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 15
Slide 41: Understanding Causality
Let’s say you’ve determined there is a real
association between our exposure and our
outcome. We believe it to be causal. That is
we’ve ruled out bias and confounding. Have we
really proven causality?
Slide 42: Surgeon General’s Guidelines for
Establishing Causality
So have you proven causality if you’ve ruled out
bias and confounding? Now the Koch-Henle
postulates are not applicable to chronic disease
epidemiology. But Doll and Hill in the 1950s were
trying to link smoking to lung cancer through a
series of case-control and cohort studies. And
these led to the development of the Surgeon General’s Guidelines for establishing
causality. These have since been through one round of revisions since the 1950s. Here
are the nine guidelines.
Slide 43: Temporal Relationship
The first guideline is whether there is a temporal
relationship between the exposure and the
disease of interest. That is the exposure must
have occurred before the disease developed.
This is easiest to establish in a prospective cohort
study in which all participants begin the study
disease free. There are couple considerations
regarding this guideline. The length of the interval between exposure and disease is very
important. For example, asbestos exposure and lung cancer. Did lung cancer follow
exposure by 3 years or 20 years? That is going to have an effect on whether this
guideline is satisfied or not.
Let’s say you have determined:
There is a real association
You believe it to be causal (ruled out confounding)
NOW have you proven CAUSALITY?
Let’s say you have determined:
There is a real association
You believe it to be causal (ruled out confounding)
NOW have you proven CAUSALITY?
1. Temporal relationship2. Strength of the association3. Dose-response relationship4. Replication of the findings5. Biologic plausibility6. Consideration of alternate explanations7. Cessation of exposure8. Consistency with other knowledge9. Specificity of the association
1. Temporal relationship2. Strength of the association3. Dose-response relationship4. Replication of the findings5. Biologic plausibility6. Consideration of alternate explanations7. Cessation of exposure8. Consistency with other knowledge9. Specificity of the association
Exposure to factor must have occurred before disease developed
Easiest to establish in a prospective cohort study
Length of interval between exposure and disease very important
e.g. asbestos and lung cancer
Lung cancer followed exposure by 3 or 20 years?
Exposure to factor must have occurred before disease developed
Easiest to establish in a prospective cohort study
Length of interval between exposure and disease very important
e.g. asbestos and lung cancer
Lung cancer followed exposure by 3 or 20 years?
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 16
Slide 44: Strength of the Association
The next guideline is the strength of the
association. The stronger the observed
association, the more likely the exposure is
causing the disease. That is, strong association is
more likely to be causal because they are
unlikely to be entirely due to bias and or
confounding. So a weak association may be
causal but it is harder to rule out bias and confounding. So here’s an example of strong
association. Relative risk of lung cancer in smokers vs. non-smokers is 9. The relative risk
in heavy vs. non-smokers is 20. These are strong associations. So it is unlikely that they
are entirely due to bias or confounding.
Slide 45: Strength of the Association
So here are three odds ratios (OR) and the
accompanying 95% confidence intervals (CI).
Which odds ratio would be more likely to infer
causation? OR#1 is 1.4 with a very narrow CI.
OR#2 is 9.8 with a very wide CI. OR#3 is 6.6 with
a fairly narrow CI. The third OR would give us
the strongest evidence towards an association,
OR of 6.6 with a fairly narrow CI. As compared with an OR of 9.8 which is stronger but
has a very wide CI.
Slide 46: Dose-Response Relationship
The next guideline is whether a dose-response
relationship exists. That is, persons who have
higher exposure should have higher risks of the
disease. For example, lung cancer death rates
rise with the number of cigarettes smoked. This
is not considered necessary for a causal
relationship but it does provide additional
evidence that a causal relationship exists. The higher the exposure, the higher the risk of
disease.
The stronger the association, the more likely the exposure is causing the disease
Example: RR of lung cancer in smokers vs. non-smokers = 9; RR of lung cancer in heavy vs. non-smokers = 20
The stronger the association, the more likely the exposure is causing the disease
Example: RR of lung cancer in smokers vs. non-smokers = 9; RR of lung cancer in heavy vs. non-smokers = 20
Which odds ratio (OR) would you be more likely to infer causation from?
OR#1: OR = 1.4 95% CI = (1.2 - 1.7)
OR#2:OR = 9.8 95% CI = (1.8 - 12.3)
OR#3:OR = 6.6 95% CI = (5.9 - 8.1)
Which odds ratio (OR) would you be more likely to infer causation from?
OR#1: OR = 1.4 95% CI = (1.2 - 1.7)
OR#2:OR = 9.8 95% CI = (1.8 - 12.3)
OR#3:OR = 6.6 95% CI = (5.9 - 8.1)
Persons who have increasingly higher exposure levels have increasingly higher risks of disease
Example: Lung cancer death rates rise with the number of cigarettes smoked
Persons who have increasingly higher exposure levels have increasingly higher risks of disease
Example: Lung cancer death rates rise with the number of cigarettes smoked
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 17
Slide 47: Age-Adjusted Mortality Rates
Here is a figure depicting a dose response
relationship between smoking and lung cancer.
This graph shows the age-adjusted mortality
rates of bronchogenic carcinoma by current
amount of smoking. We have mortality rate by
100,000 person-years on the y-axis and the
amount of smoking on the x-axis. As you can
see, the mortality rate per 100,000 among the ‘never’ smokers was 3.4. And it steadily
increases with the number of packs smoked a day up to ‘2+ packs/day’ where the
mortality rate is 217.3. Clearly a dose response.
Slide 48: Replication of Findings
The next guideline is replication of findings. Your
observed association linking exposure to disease
should be observed repeatedly in different
persons and populations, places, times, and
circumstances, using different study designs, and
different investigators. For example, smoking
has been associated with lung cancer in dozens
of retrospective and prospective studies all around the world using different study
designs and among different populations.
Slide 49: Biologic Plausibility
The next guideline is biologic plausibility. There
must be a biological or social model to explain
your observed association. That is, your
observed association should not conflict with the
current knowledge or natural history and biology
of the disease. For example cigarettes contain
many carcinogenic substances. However, many
epidemiologic studies have identified this causal relationship before the biological
mechanisms were identified.
The association is observed repeatedly in different persons, places, times, and circumstances
Replicating the association in different samples, with different study designs, and different investigators gives evidence of causation
Example: Smoking has been associated with lung cancer in dozens of retrospective and prospective studies
The association is observed repeatedly in different persons, places, times, and circumstances
Replicating the association in different samples, with different study designs, and different investigators gives evidence of causation
Example: Smoking has been associated with lung cancer in dozens of retrospective and prospective studies
Biological or social model exists to explain the association Does not conflict with current knowledge of natural history
and biology of disease
Example: Cigarettes contain many carcinogenic substances
Many epidemiologic studies have identified causal relationships before biological mechanisms were identified
Biological or social model exists to explain the association Does not conflict with current knowledge of natural history
and biology of disease
Example: Cigarettes contain many carcinogenic substances
Many epidemiologic studies have identified causal relationships before biological mechanisms were identified
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 18
Slide 50: Considerations of Alternate
Explanations
Next, the consideration of alternate
explanations. So did investigators rule out bias
and confounding? And these are alternate
explanations. So did investigators consider the
associations between smoking and coffee
consumption and pancreatic cancer? This gets
back to our earlier discussion of bias and confounding. These must be ruled out in order
to determine whether a causal relationship exists.
Slide 51: Cessation of Exposure
Next, cessation of exposure. The risk of
developing the disease should decline when the
exposure has been reduced or eliminated.
However, in certain cases the damage may be
irreversible. For example, emphysema is not
reversed with the cessation of smoking, but its
progression can be reduced. In the famous
example of John Snow investigating a risk factor of cholera being contaminated water, he
removed the pump handle at the Broad St. pump and found that the risk of disease
decreased after the handle was removed.
Slide 52: Consistency With Other Knowledge
Next, is your observed association consistent
with other knowledge? So if the relationship is
causal, your relationship should be consistent
with other data. So if lung cancer incidence
increased as cigarette use was on the decline,
you need to explain how this was consistent with
a causal relationship because this shouldn’t be
the case.
Did the investigators consider bias and confounding?
Investigators must consider other possible explanations
Example: Did the investigators consider the associations between smoking, coffee consumption and pancreatic cancer?
Did the investigators consider bias and confounding?
Investigators must consider other possible explanations
Example: Did the investigators consider the associations between smoking, coffee consumption and pancreatic cancer?
Risk of disease should decline when exposure to factor is reduced or eliminated
In certain cases, the damage may be irreversible
Example: Emphysema is not reversed with the cessation of smoking, but its progression is reduced
Risk of disease should decline when exposure to factor is reduced or eliminated
In certain cases, the damage may be irreversible
Example: Emphysema is not reversed with the cessation of smoking, but its progression is reduced
If a relationship is causal, the findings should be consistent with other data
If lung cancer incidence increased as cigarette use was on the decline, need to explain how this was consistent with a causal relationship
If a relationship is causal, the findings should be consistent with other data
If lung cancer incidence increased as cigarette use was on the decline, need to explain how this was consistent with a causal relationship
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 19
Slide 53: Specificity of the Association
Next, specificity of the association. A single
exposure should cause a single disease. This is a
very narrow guideline. For example, smoking is
associated with lung cancer as well as many
other diseases. And lung cancer results from
smoking as well as many other exposures. So
this guideline is very difficult to satisfy and it has
its roots back in the Koch-Henle postulates. However, when this is present, like other
guidelines, it provides additional support for a causal relationship. However, if this
guideline cannot be satisfied, it does not preclude a causal relationship.
Slide 54: Uses of Surgeon General’s Guidelines for Establishing Causality How can we use the Surgeon General’s Guidelines for Establishing Causality? We can use these guidelines to remember distinctions between association and causation in epidemiologic research. It needs to be in the back of every investigator’s mind that if association is established it does not confirm
causation. Next, these guidelines are useful when critically evaluating or reading epidemiologic studies. These guidelines can also be useful for designing epidemiologic studies. Different study design options or different methods within a study design can be used so that these guidelines can be satisfied. For example a dose-response or a temporal relationship between your exposure and your outcome. Next these guidelines can be useful when interpreting the results of your own study. And finally, overall these guidelines are not meant to be rigid. If one guideline is not satisfied that does not preclude a causal relationship. These guidelines are used to essentially build a case for exposure causing a disease. One study is not going to prove causality. It’s the body of work or body of knowledge about an exposure and an outcome.
Remembering distinctions between association and causation in epidemiologic research
Critically reading epidemiologic studies
Designing epidemiologic studies
Interpreting the results of your own study
Remembering distinctions between association and causation in epidemiologic research
Critically reading epidemiologic studies
Designing epidemiologic studies
Interpreting the results of your own study
A single exposure should cause a single disease
Example: Smoking is associated with lung cancer as well as many other diseases Lung cancer results from smoking as well as other exposures
When present, provides additional support for causal inference
When absent, does not preclude a causal relationship
A single exposure should cause a single disease
Example: Smoking is associated with lung cancer as well as many other diseases Lung cancer results from smoking as well as other exposures
When present, provides additional support for causal inference
When absent, does not preclude a causal relationship
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 20
Slide 55: Does HIV Cause AIDS?
Let’s look at an example: Does HIV cause AIDS?
The vast majority of scientists believe that HIV
causes AIDS however; there is a small group that
believes AIDS is a behavioral rather than an
infectious disease. They believe that AIDS is
caused by the use of recreational drugs and the
treatment drugs offered in the U.S. and Europe,
and by malnutrition in Africa.
Slide 56: Does HIV Cause AIDS?
Let’s look back at the Koch-Henle Postulates.
Regular and exclusive? Yes. Gallo et al. routinely
found HIV in people with AIDS and failed to find
HIV among people who either lacked AIDS
symptoms or the AIDS-associated risk factors.
Regarding the experimental model with the 3rd
postulate, 3 lab workers were accidentally
infected in the early 1990s with purely molecularly cloned HIV. One developed
pneumonia (an AIDS-defining disease) before starting antiretroviral therapy. That would
negate the argument that the therapy is causing the disease.
Slide 57: Does HIV Cause AIDS?
Let’s look at the guidelines. There have been
numerous epidemiological studies which have
established a temporal relationship. That is
infection with the virus precedes the
development of AIDS-related symptoms. There
is a strong association between the virus and the
symptoms. There is a dose-response and
replication of findings. This is biologically plausible. Cessation of exposure: there is a
decrease in deaths associated with AIDS after the advent of antiretroviral therapy. This is
a very specific association. And it’s consistent with other knowledge.
Regular and exclusive
Gallo et al. routinely found HIV in people with AIDS symptoms and failed to find HIV among people who either lacked AIDS symptoms or AIDS-associated risk factors
Experimental model
3 lab workers accidentally infected in early 1990s with purely molecularly cloned strain of HIV
One developed pneumonia (AIDS-defining disease) before started antiretroviral therapy
Regular and exclusive
Gallo et al. routinely found HIV in people with AIDS symptoms and failed to find HIV among people who either lacked AIDS symptoms or AIDS-associated risk factors
Experimental model
3 lab workers accidentally infected in early 1990s with purely molecularly cloned strain of HIV
One developed pneumonia (AIDS-defining disease) before started antiretroviral therapy
Epidemiologic studies have established:
Temporal relationship
Strong association
Dose response
Replication of findings
Biologic plausibility
Cessation of exposure (decrease in deaths after antiretroviral therapy)
Specificity
Consistency with other knowledge
Epidemiologic studies have established:
Temporal relationship
Strong association
Dose response
Replication of findings
Biologic plausibility
Cessation of exposure (decrease in deaths after antiretroviral therapy)
Specificity
Consistency with other knowledge
Majority of scientists believe HIV causes AIDS
Small group believes AIDS is a behavioral rather than an infectious disease
AIDS is caused by use of recreational drugs and antiretroviral drugs in the U.S. and Europe, and by malnutrition in Africa
Majority of scientists believe HIV causes AIDS
Small group believes AIDS is a behavioral rather than an infectious disease
AIDS is caused by use of recreational drugs and antiretroviral drugs in the U.S. and Europe, and by malnutrition in Africa
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 21
Slide 58: Causation
A few final words on causation. Remember
associations are observed, and causation is
inferred. So it’s important to remember that
these guidelines provide evidence for causal
relationships. They don’t really prove a causal
relationship. Next, all of the evidence must be
considered and the criteria weighed against each
other to infer that there is a causal relationship.
Slide 59: Summary
In summary, we discussed three aspects of an
association. Whether the association is biased:
was there a systematic error that resulted in a
flawed or incorrect estimate of association? We
talked about selection and information bias.
Next we talked about whether this observed
association was confounded, that there was an
alternate explanation, which resulted in this observed association. We talked about
controlling confounding in the design and analysis phases. Lastly, if bias and confounding
can be ruled out, is the observed association causal? It’s important to remember
causation must be inferred, whereas associations are observed. Thank you and that
concludes Module 2: Issues of Interpretation in Epidemiologic Studies.
Center for Public Health Continuing EducationUniversity at Albany School of Public Health
Department of Community & Family MedicineDuke University School of Medicine
Associations are observedCausation is inferred
It is important to remember that these criteria provide evidence for causal relationships
All of the evidence must be considered and the criteria weighed against each other to infer the causal relationship
Associations are observedCausation is inferred
It is important to remember that these criteria provide evidence for causal relationships
All of the evidence must be considered and the criteria weighed against each other to infer the causal relationship
Bias is a systematic error that results in an incorrect estimate of association
Selection and information bias
Confounding is an alternate explanation which can be controlled
Design and analysis phases
Causation must be inferred
Bias is a systematic error that results in an incorrect estimate of association
Selection and information bias
Confounding is an alternate explanation which can be controlled
Design and analysis phases
Causation must be inferred
Module 2: Fundamentals of Epidemiology – Issues of Interpretation
Transcript
Page 22
Mike Barry, CAELorrie Basnight, MDNancy Bennett, MD, MSRuth Gaare Bernheim, JD, MPHAmber Berrian, MPHJames Cawley, MPH, PA-CJack Dillenberg, DDS, MPHKristine Gebbie, RN, DrPHAsim Jani, MD, MPH, FACP
Denise Koo, MD, MPHSuzanne Lazorick, MD, MPHRika Maeshiro, MD, MPHDan Mareck, MDSteve McCurdy, MD, MPHSusan M. Meyer, PhDSallie Rixey, MD, MEdNawraz Shawir, MBBS
Sharon Hull, MD, MPHPresident
Allison L. LewisExecutive Director
O. Kent Nordvig, MEd
Project Representative